New ideas for the experimental protocol
18 November 2001
Francis to Martie

And I had an idea: what if we had a poster of a real-life person, and compared how subjects respond to that? The idea is to track how people make use of images according to how they frame them. When you look at the Mona Lisa, you can legitimately ask questions about what she's thinking and feeling -- it's somehow part of the game of looking at art. Imagine instead looking at a photograph of a past colleague -- the kind of large posters departments sometimes have. You don't look closely at that and try to reconstruct what precisely the person was thinking at the time -- first of all, it's completely irrelevant (who cares), secondly, you just can't know. Maybe her kid got measles that morning, or her worst enemy got promoted, and it shows up as a certain look in her face -- we can't know and we don't care. In the executive case, in real life, we have to have the story to make sense of something like that, and we'd need to be part of the story in some way for it to be relevant to us. So the idea is that there is a tacit set of rules that apply to art that don't apply to real-life presentations -- and this is not a question exactly of realism or not -- a hyperrealistic painting would still invite a different kind of looking than a photograph of the same scene.

So what I'd like to do is to try to get to how subjects may differentially immerse in artistic images in ways they don't tend to immerse in photographs of real people taken for historical reasons. It occurred to me we could use photographs from online dating sites -- so we would compare paintings of physically attractive people with photographs of physically attractive people. In the former case, we anticipate imaginative immersion in the absence of any real information about the facts -- in fictive character or even if he was real he's now long since dead. The immersion should in my hypothesis take place without primarily trying to relate the depicted person to the viewer's real self (executive agent): people should as it were enter the fictive scenario and forget about themselves.

So the question is, in the second case, when subjects see a photograph of someone advertising on a dating site, will they respond differently or the same? If they were really looking for someone to date, they would of course be in the executive mode, and they would try to be as realistic as possible about the person depicted. They would make inferences, for instance, about the composition of the picture itself as a communicative act, intentional or not -- that is to say, they would perhaps tend to think that a messy background might suggest the person really is very messy, if she couldn't be bothered to tidy her room even when she was getting a photograph taken to attract a mate. In the executive mode, they would want to have ways of checking up on their inferences, and willing to change them when they got new information: they could find out later, say, that it's the agency that takes the pictures, and the room is at the agency, not at her house; the mess wasn't hers and she's in fact irritatingly anal. When you look at a real photograph, you realize these types of possibilities and make allowances for them.

On the other hand, if they were just looking at a photograph of a woman trying to attract a mate, where they have no information about where the photograph is from and no possibility of acting on the information they derive, it becomes more difficult to characterize exactly how they may make use of it. In short, they should be very sensitive to the experimental situation: if they are paid to run a simulation of seeing that person for real and evaluating her as a mate, they will do that -- but is that executive or organizational? It isn't obviously either one: they are not trying to achieve something in actuality with regard to the woman depicted, nor are they immersing imagnatively in the organizational mode just for fun. The simplest way to think about it may be to say that it is simply ecologically anomalous: given a financial or other executive incentive, people are simply able to tap into their minds in a way that isn't functional in itself. Of course, the capacity is there because it made a difference in survival in some set of proper domains -- say, for pretend play or for planning -- but psychology experiments fit neither of these domains; they are an artificial usage of evolved capacities. We therefore wouldn't expect the experiments to reveal the full set of interpretive algoriths proper to either adaptive domain.

This means that if we're going to get useful and clean results on the different inference rules used in organizational vs executive mode interactions, we need to be very careful about how the task is framed. You can't simply say, "Evaluate this person as a candidate for dating" and expect the person to enter the executive mode -- the subject won't really have any reason to think the candidate is really available, she may already have a mate, she may in principle dislike dating agencies, etc. The point is that you won't get a clean executive mode unless the person uses the information to act in his real life in ways that make a difference that relates to his survival, if only in a small way. This is not something we can expect to manipulate successfully in the lab.

So if we would like to specifically investigate the differences between the executive and the organizational mode responses, we should at least attempt to provide a set of cues designed to move subjects closer to the orginal adaptive domains. In the case of photographs of real people, we could create well-defined fictive scenarios: imagine you're looking for a mate, or imagine you're out to hire someone. Tell me how you interpret the picture. Or to elicit the organizational mode: encourage slow and relaxed imaginative immersion, discourage executive mode perspectives. Our results won't be clean, but if we get an effect, the real phenomenon is likely to be stronger.

(As an aside, there appear to be a whole set of phenomena in the borderlands between executive and organizational -- phenomena like television news that aren't strictly speaking entertainment, but that also don't require you to do a thing -- in a sense, that is information that you don't really need and that you can therefore afford to treat as affordances for enjoyable imaginative immersion.)

So that's the suggestion: we use some photographs of real people, such as posters of past UCLA professors and perhaps film stars if we can get hold of them, and also online photographs from dating services, and we place the subjects in a set of imaginative situations. These imaginative situations are not in the organizational mode or the executive mode; they are simply experimental constructs for allowing us to get access to the subject's cognitive operations, more or less closely approximating the operations of these processes in their natural settings. We create a set of these imaginative situations and apply them across the board to art and real-life photographs. The point is that it's the framing that will trigger the cognitive operations we're interested in. The photographs and posters themselves will not suffice.

Proposed framing:

• "You're a young man/woman in search of a mate. Evaluate the following candidates."
  
  
• "You're on vacation and have no duties. In an art museum, you find the following paintings and photographs."

Those are the obvious ones. Of course, it's tempting to expand with a couple of contrasting executive frames:

• "You run an accounting firm and need to hire a secretary. Please evaluate the following applicants."
  
  
• "You're a detective in the homicide division of a police department. A girl has been murdered. Consider the following list of suspects."

The point here -- apart from the fact that it would be hilarious: would the Mona Lisa make a good accountant? did she just kill her maid? -- is that we want to bring out the distinctiveness of an artistic/aesthetic treatment of something. Here, the theory can be more cleanly formulated: any stimulus should lend itself in some way to an aesthetic experience, but some objects are better than others, in part because they are deliberately ambiguous. We should also see if people in fact are more willing to speculate about mental states if the frame is a museum, or perhaps differ in their choice of speculatively attributed mental states.

Let me know what you think about this scenario.

Let me add some thoughts on the issue of prediction. In the organizational-mode hypothesis, people use affordances in objects to construct imaginative scenarios that allow they to practice situations that in the EEA were important for survival. This practice, according to the model, should follow a developmentally determined implicit curriculum; while we also predict situational sensitivity, this should be primarily to local cultural conditions (relating to informational constituencies of local knowledge). We don't predict a strong effect from local variations in personal situation -- for instance, we would not predict that women without a boyfriend would find romantic stories a great deal more attractive than women with boyfriends (actually, this would be fun to test -- perhaps we can pull that out of the data from this project?) This contrasts with Freud's idea that fiction is a form of personal wish fulfillment. With respect to the gender bias you find in potentially flirtatious interactions, the organizational-mode hypothesis has this to say:

1. Insofar as finding a mate is a developmentally bound task, we would definitely expect that interest in fictive scenarios that involve problem-solving related to mating will correlate with age and sex, in ways that we could discuss in more detail.
2. However, since art according to this model consist of affordances that are used in facultative ways by the viewer, we do not predict that the viewers of art will see the depictions of physically attractive people as people who are attempting to communicate with them, or as potential mates. It is therefore not clear that we would expect a gender-based interpretive bias. We should discuss this in more detail and thrash out what we expect.
3. If we use a mate-selection pretend frame, on the other hand, we should get the bias.

In light of this, using deliberately engineered pretend frames may get us more interesting data.

Finally, I have some ideas about how we can start the session, with respect to tracking the subjects' assignment of representational to the posters. We could simply let the person get comfortable, after having looked around, and ask, Could you list the items you see in this room? At this stage, we'd be more likely to get a description of the posters as objects, fully differentiated from the real people in the room.

Addition 19 November 2001:

Summary points:

1. The organizational mode hypothesis (OMH) predicts developmentally appropriate entertainment preferences
2. Because of the absence of social constraints or real-life consequences, sex-specific biases in communication should be greater in the organizational mode (OM)
3. In fact, entertainment preferences appear to vary significantly by sex
4. So wouldn't we predict an increased bias by sex in artistic preferences and in the cognitive processes evoked by artistic objects depicting sexually attrative individuals?

Discussion:

1. The observed sex-related differences in entertainment preference (what do we have in terms of good studies on this?) may be addressing only movies and narrative fiction. In those cases, the fictive scenario provides a sex-appropriate fictional stand-in, which can function as the vehicle for the viewer's virtual agent. In these cases, the OMH would firmly predict large sex differences, as the developmental tasks to be trained differ by sex.
2. In the case of portraits of one individual, it is not possible to view the fictive individual depicted as the object of another fictive person's affection, since there is no other fictive person depicted. This means that we would have to have the executive agent relate to the fictive depiction; the idea that this would typically happen is not a prediction of the OHM.
3. We could easily create frames where there is a fictive agent that could constitute a story -- for instance, we could ask subjects to imagine they are a courtier flirting with the Mona Lisa, or that they are the Mona Lisa flirting with a hidden courtier -- say, a fictive viewer of the painting who was the Mona Lisa's contemporary.
4. However, this doesn't really get to the question of Mona Lisa's smile. We may just have to leave it as an empirical question whether males are more likely to see the smile as flirtatious or not -- I still dont' see how we can generate a hypothesis consistent with the OMH at this point. There's probably something I'm missing.

Let me know when you've had a chance to consider the idea of manipulating the situational frame to get out the differences between the OM and the executive mode (EM). It would be interesting if, say, we found a bias effect in EM but not in OM. Very interesting, actually.